Multiprotocol decapsulation/encapsulation control structure and packet protocol conversion method

BACKGROUND

0001This invention relates to forwarding network packets between network domains.

0002Packets are routed through a series of router devices, each of which stores and forwards packets on its way from a source to a destination. For example, a packet may start out as an Internet packet, be forwarded over an ATM (asynchronous transfer mode path) and then back to Ethernet onto a corporate network to its final intended recipient. As the network passes through these network domains, various header encapsulations may be added to or removed from the packet. Some connections use point-to-point protocol (PPP) whereas others use multiprotocol label switching MPLS, layer to tunneling protocol LTTP, ATM and so forth.

DESCRIPTION OF DRAWINGS

0003<figref idref="DRAWINGS">FIG. 1</figref> is a block diagram of a communication system employing a hardware based multithreaded processor.

0004<figref idref="DRAWINGS">FIGS. 2-1</figref> to <b>2</b>-<b>4</b> are a detailed block diagram of a hardware based multithreaded processor of <figref idref="DRAWINGS">FIG. 1</figref>.

0005<figref idref="DRAWINGS">FIG. 3</figref> is a block diagram depicting a functional arrangement of the multithreaded processor of <figref idref="DRAWINGS">FIG. 2</figref>.

0006<figref idref="DRAWINGS">FIG. 4</figref> is a block diagram depicting data structures in memory used for the processor of <figref idref="DRAWINGS">FIG. 1</figref>.

0007<figref idref="DRAWINGS">FIG. 5</figref> is a block diagram depicting formats for forwarding tables used in the tables of <figref idref="DRAWINGS">FIG. 4</figref>.

0008<figref idref="DRAWINGS">FIG. 6</figref> is a flow chart depicting a generic packet forwarding process.

0009<figref idref="DRAWINGS">FIG. 7</figref> is a flow chart depicting an alternative aspect of the packet forwarding process of <figref idref="DRAWINGS">FIG. 6</figref>.

DETAILED DESCRIPTION

0010Referring to <figref idref="DRAWINGS">FIG. 1</figref>, a communication system <b>10</b> includes a parallel, hardware-based multithreaded processor <b>12</b>. The hardware-based multithreaded processor <b>12</b> is coupled to a bus such as a PCI bus <b>14</b>, a memory system <b>16</b> and a second bus <b>18</b>. The system <b>10</b> is especially useful for tasks that can be broken into parallel subtasks or functions. Specifically hardware-based multithreaded processor <b>12</b> is useful for tasks that are bandwidth oriented rather than latency oriented. The hardware-based multithreaded processor <b>12</b> has multiple microengines <b>22</b> each can be simultaneously active and work on multiple hardware controlled threads that independently work on a task.

0011The hardware-based multithreaded processor <b>12</b> also includes a central controller <b>20</b> that assists in loading microcode control for other resources of the hardware-based multithreaded processor <b>12</b> and performs other general purpose computer type functions such as handling protocols, exceptions, extra support for packet processing where the microengines pass the packets off for more detailed processing such as in boundary conditions. In one embodiment, the processor <b>20</b> is a Strong Arm® based architecture. The general purpose microprocessor <b>20</b> has an operating system. Through the operating system the processor <b>20</b> can call functions to operate on microengines <b>22</b><i>a</i>–<b>22</b><i>f</i>. The processor <b>20</b> can use any supported operating system preferably a real time operating system such as, MicrosoftNT real-time, VXWorks.

0012The hardware-based multithreaded processor <b>12</b> also includes a plurality of function microengines <b>22</b><i>a</i>–<b>22</b><i>f</i>. Functional microengines (microengines) <b>22</b><i>a</i>–<b>22</b><i>f </i>each maintain a plurality of program counters in hardware and states associated with the program counters. Effectively, a corresponding plurality of sets of threads can be simultaneously active on each of the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>while only one is actually operating at any one time.

0013In one embodiment, there are six microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>as shown. Each microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>has capabilities for processing four hardware threads. The six microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>operate with shared resources including memory system <b>16</b> and bus interfaces <b>24</b> and <b>28</b>. The memory system <b>16</b> includes a Synchronous Dynamic Random Access Memory (SDRAM) controller <b>26</b><i>a </i>and a Static Random Access Memory (SRAM) controller <b>26</b><i>b</i>. SDRAM memory <b>16</b><i>a </i>and SDRAM controller <b>26</b><i>a </i>are typically used for processing large volumes of data, e.g., processing of network payloads from network packets. The SRAM controller <b>26</b><i>b </i>and SRAM memory <b>16</b><i>b </i>are used in a networking implementation for low latency, fast access tasks, e.g., accessing look-up tables, memory for the core processor <b>20</b>, and so forth.

0014The six microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>access either the SDRAM <b>16</b><i>a </i>or SRAM <b>16</b><i>b </i>based on characteristics of the data. Thus, low latency, low bandwidth data is stored in and fetched from SRAM, whereas higher bandwidth data for which latency is not as important, is stored in and fetched from SDRAM. The microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>can execute memory reference instructions to either the SDRAM controller <b>26</b><i>a </i>or SRAM controller <b>16</b><i>b. </i>

0015Advantages of hardware multithreading can be explained by SRAM or SDRAM memory accesses. As an example, an SRAM access requested by a Thread_—0, from a microengine will cause the SRAM controller <b>26</b><i>b </i>to initiate an access to the SRAM memory <b>16</b><i>b</i>. The SRAM controller controls arbitration for the SRAM bus, accesses the SRAM <b>16</b><i>b</i>, fetches the data from the SRAM <b>16</b><i>b</i>, and returns data to a requesting microengine <b>22</b><i>a</i>–<b>22</b><i>b</i>. During an SRAM access, if the microengine e.g., <b>22</b><i>a </i>had only a single thread that could operate, that microengine would be dormant until data was returned from the SRAM. By employing hardware context swapping within each of the microengines <b>22</b><i>a</i>–<b>22</b><i>f</i>, the hardware context swapping enables other contexts with unique program counters to execute in that same microengine. Thus, another thread e.g., Thread_—1 can function while the first thread, e.g., Thread_—0, is awaiting the read data to return. During execution, Thread_—1 may access the SDRAM memory <b>16</b><i>a</i>. While Thread_—1 operates on the SDRAM unit, and Thread_—0 is operating on the SRAM unit, a new thread, e.g., Thread_—2 can now operate in the microengine <b>22</b><i>a</i>. Thread_—2 can operate for a certain amount of time until it needs to access memory or perform some other long latency operation, such as making an access to a bus interface. Therefore, simultaneously, the processor <b>12</b> can have a bus operation, SRAM operation and SDRAM operation all being completed or operated upon by one microengine <b>22</b><i>a </i>and have one more thread available to process more work in the data path.

0016The hardware context swapping also synchronizes completion of tasks. For example, two threads could hit the same shared resource e.g., SRAM. Each one of these separate functional units, e.g., the FBUS interface <b>28</b>, the SRAM controller <b>26</b><i>a</i>, and the SDRAM controller <b>26</b><i>b</i>, when they complete a requested task from one of the microengine thread contexts reports back a flag signaling completion of an operation. When the flag is received by the microengine, the microengine can determine which thread to turn on.

0017One example of an application for the hardware-based multithreaded processor <b>12</b> is as a network processor. As a network processor, the hardware-based multithreaded processor <b>12</b> interfaces to network devices such as a media access controller device e.g., a 10/100BaseT Octal MAC <b>13</b><i>a </i>or a Gigabit Ethernet device <b>13</b><i>b</i>. In general the network process can interface to any type of communication device or interface that receives/sends large amounts of data. Communication system <b>10</b> functioning in a networking application could receive a plurality of network packets from the devices <b>13</b><i>a</i>, <b>13</b><i>b </i>and process those packets in a parallel manner. With the hardware-based multithreaded processor <b>12</b>, each network packet can be independently processed. Another example for use of processor <b>12</b> is a print engine for a postscript processor or as a processor for a storage subsystem, i.e., RAID disk storage. A further use is as a matching engine. In the securities industry for example, the advent of electronic trading requires the use of electronic matching engines to match orders between buyers and sellers. These and other parallel types of tasks can be accomplished on the system <b>10</b>.

0018The processor <b>12</b> includes a bus interface <b>28</b> that couples the processor to the second bus <b>18</b>. Bus interface <b>28</b> in one embodiment couples the processor <b>12</b> to the so-called FBUS <b>18</b> (FIFO bus). The FBUS interface <b>28</b> is responsible for controlling and interfacing the processor <b>12</b> to the FBUS <b>18</b>. The FBUS <b>18</b> is a 64-bit wide FIFO bus, which is currently gaining acceptance as the best bus for Media Access Controller (MAC) devices.

0019The processor <b>12</b> includes a second interface e.g., a PCI bus interface <b>24</b> that couples other system components that reside on the PCI <b>14</b> bus to the processor <b>12</b>. The PCI bus interface <b>24</b>, provides a high speed data path <b>24</b><i>a </i>to memory <b>16</b> e.g., the SDRAM memory <b>16</b><i>a</i>. Through that path data can be moved quickly from the SDRAM <b>16</b><i>a </i>through the PCI bus <b>14</b>, via direct memory access (DMA) transfers. Additionally, the PCI bus interface <b>24</b> supports target and master operations. Target operations are operations where slave devices on bus <b>14</b> access SDRAMs through reads and writes that are serviced as a slave to target operation. In master operations, the processor core <b>20</b> sends data directly to or receives data directly from the PCI interface <b>24</b>.

0020Each of the functional units are coupled to one or more internal buses. The processor includes an AMBA bus that couples the processor core <b>20</b> to the memory controller <b>26</b><i>a</i>, <b>26</b><i>c </i>and to an AMBA translator <b>30</b> described below. The processor also includes a private bus <b>34</b> that couples the microengine units to SRAM controller <b>26</b><i>b</i>, AMBA translator <b>30</b> and FBUS interface <b>28</b>. A memory bus <b>38</b> couples the memory controller <b>26</b><i>a</i>, <b>26</b><i>b </i>to the bus interfaces <b>24</b> and <b>28</b> and memory system <b>16</b> including flashrom <b>16</b><i>c </i>used for boot operations and so forth.

0021Referring to <figref idref="DRAWINGS">FIGS. 2-1</figref> to <b>2</b>-<b>4</b>, each of the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>includes an arbiter that examines flags to determine the available threads to be operated upon. Any thread from any of the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>can access the SDRAM controller <b>26</b><i>a</i>, SDRAM controller <b>26</b><i>b </i>or FBUS interface <b>28</b>. The memory controllers <b>26</b><i>a </i>and <b>26</b><i>b </i>each include a plurality of queues to store outstanding memory reference requests. The queues either maintain order of memory references or arrange memory references to optimize memory bandwidth. For example, if a thread_—0 has no dependencies or relationship to a thread_—1, there is no reason that thread 1 and 0 cannot complete their memory references to the SRAM unit out of order. The microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>issue memory reference requests to the memory controllers <b>26</b><i>a </i>and <b>26</b><i>b</i>. The microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>flood the memory subsystems <b>26</b><i>a </i>and <b>26</b><i>b </i>with enough memory reference operations such that the memory subsystems <b>26</b><i>a </i>and <b>26</b><i>b </i>become the bottleneck for processor <b>12</b> operation.

0022If the memory subsystem <b>16</b> is flooded with memory requests that are independent in nature, the processor <b>12</b> can perform memory reference sorting. Memory reference sorting improves achievable memory bandwidth. Memory reference sorting, as described below, reduces dead time or a bubble that occurs with accesses to SRAM. With memory references to SRAM, switching current direction on signal lines between reads and writes produces a bubble or a dead time waiting for current to settle on conductors coupling the SRAM <b>16</b><i>b </i>to the SRAM controller <b>26</b><i>b. </i>

0023That is, the drivers that drive current on the bus need to settle out prior to changing states. Thus, repetitive cycles of a read followed by a write can degrade peak bandwidth. Memory reference sorting allows the processor <b>12</b> to organize references to memory such that long strings of reads can be followed by long strings of writes. This can be used to minimize dead time in the pipeline to effectively achieve closer to maximum available bandwidth. Reference sorting helps maintain parallel hardware context threads. On the SDRAM, reference sorting allows hiding of pre-charges from one bank to another bank. Specifically, if the memory system <b>16</b><i>b </i>is organized into an odd bank and an even bank, while the processor is operating on the odd bank, the memory controller can start precharging the even bank. Precharging is possible if memory references alternate between odd and even banks. By ordering memory references to alternate accesses to opposite banks, the processor <b>12</b> improves SDRAM bandwidth.

0024The FBUS interface <b>28</b> supports Transmit and Receive flags for each port that a MAC device supports, along with an Interrupt flag indicating when service is warranted. The FBUS interface <b>28</b> also includes a controller <b>28</b><i>a </i>that performs header processing of incoming packets from the FBUS <b>18</b>. The controller <b>28</b><i>a </i>extracts the packet headers and performs a microprogrammable source/destination/protocol hashed lookup (used for address smoothing) in SRAM. If the hash does not successfully resolve, the packet header is sent to the processor core <b>20</b> for additional processing. The FBUS interface <b>28</b> supports the following internal data transactions:

0025<tables id="TABLE-US-00001" num="00001"><table frame="none" colsep="0" rowsep="0"><tgroup align="left" colsep="0" rowsep="0" cols="4"><colspec colname="offset" colwidth="14pt" align="left" /><colspec colname="1" colwidth="49pt" align="left" /><colspec colname="2" colwidth="63pt" align="left" /><colspec colname="3" colwidth="91pt" align="left" /><thead><row><entry /><entry namest="offset" nameend="3" align="center" rowsep="1" /></row></thead><tbody valign="top"><row><entry /><entry>FBUS unit</entry><entry>(via AMBA bus)</entry><entry>to/from processor Core.</entry></row><row><entry /><entry>FBUS unit</entry><entry>(via private bus)</entry><entry>to/from SRAM Unit.</entry></row><row><entry /><entry>FBUS unit</entry><entry>(via Mbus)</entry><entry>to/from SDRAM.</entry></row><row><entry /><entry namest="offset" nameend="3" align="center" rowsep="1" /></row></tbody></tgroup></table></tables>

0026The FBUS <b>18</b> is a standard industry bus and includes a data bus, e.g., 64 bits wide and sideband control for address and read/write control. The FBUS interface <b>28</b> provides the ability to input large amounts of data using a series of input and output FIFO's <b>29</b><i>a</i>–<b>29</b><i>b</i>. From the FIFOs <b>29</b><i>a</i>–<b>29</b><i>b</i>, the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>fetch data from or command the SDRAM controller <b>26</b><i>a </i>to move data from a receive FIFO in which data has come from a device on bus <b>18</b>, into the FBUS interface <b>28</b>. The data can be sent through memory controller <b>26</b><i>a </i>to SDRAM memory <b>16</b><i>a</i>, via a direct memory access. Similarly, the microengines can move data from the SDRAM <b>26</b><i>a </i>to interface <b>28</b>, out to FBUS <b>18</b>, via the FBUS interface <b>28</b>.

0027Data functions are distributed amongst the microengines. Connectivity to the SRAM <b>26</b><i>a</i>, SDRAM <b>26</b><i>b </i>and FBUS <b>28</b> is via command requests. A command request can be a memory request or a FBUS request. For example, a command request can move data from a register located in a microengine <b>22</b><i>a </i>to a shared resource, e.g., an SDRAM location, SRAM location, flash memory or some MAC address. The commands are sent out to each of the functional units and the shared resources. However, the shared resources do not need to maintain local buffering of the data. Rather, the shared resources access distributed data located inside of the microengines. This enables microengines <b>22</b><i>a</i>–<b>22</b><i>f</i>, to have local access to data rather than arbitrating for access on a bus and risk contention for the bus. With this feature, there is a 0 cycle stall for waiting for data internal to the microengines <b>22</b><i>a</i>–<b>22</b><i>f. </i>

0028The data buses, e.g., AMBA bus <b>30</b>, SRAM bus <b>34</b> and SDRAM bus <b>38</b> coupling these shared resources, e.g., memory controllers <b>26</b><i>a </i>and <b>26</b><i>b </i>are of sufficient bandwidth such that there are no internal bottlenecks. Thus, in order to avoid bottlenecks, the processor <b>12</b> has an bandwidth requirement where each of the functional units is provided with at least twice the maximum bandwidth of the internal buses. As an example, the SDRAM can run a 64 bit wide bus at 83 MHz. The SRAM data bus could have separate read and write buses, e.g., could be a read bus of 32 bits wide running at 166 MHz and a write bus of 32 bits wide at 166 MHz. That is, in essence, 64 bits running at 166 MHz which is effectively twice the bandwidth of the SDRAM.

0029The core processor <b>20</b> also can access the shared resources. The core processor <b>20</b> has a direct communication to the SDRAM controller <b>26</b><i>a </i>to the bus interface <b>24</b> and to SRAM controller <b>26</b><i>b </i>via bus <b>32</b>. However, to access the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>and transfer registers located at any of the microengines <b>22</b><i>a</i>–<b>22</b><i>f</i>, the core processor <b>20</b> access the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>via the AMBA Translator <b>30</b> over bus <b>34</b>. The AMBA translator <b>30</b> can physically reside in the FBUS interface <b>28</b>, but logically is distinct. The AMBA Translator <b>30</b> performs an address translation between FBUS microengine transfer register locations and core processor addresses (i.e., AMBA bus) so that the core processor <b>20</b> can access registers belonging to the microengines <b>22</b><i>a</i>–<b>22</b><i>c. </i>

0030The processor core <b>20</b> includes a RISC core <b>50</b> implemented in a five stage pipeline performing a single cycle shift of one operand or two operands in a single cycle, provides multiplication support and 32 bit barrel shift support. This RISC core <b>50</b> is a standard Strong Arm® architecture but it is implemented with a five stage pipeline for performance reasons. The processor core <b>20</b> also includes a 16 kilobyte instruction cache <b>52</b>, an 8 kilobyte data cache <b>54</b> and a prefetch stream buffer <b>56</b>. The core processor <b>20</b> performs arithmetic operations in parallel with memory writes and instruction fetches. The core processor <b>20</b> interfaces with other functional units via the ARM defined AMBA bus. The AMBA bus is a 32-bit bi-directional bus <b>32</b>.

0031Referring to <figref idref="DRAWINGS">FIG. 3</figref>, the multiprocessor <b>12</b> is shown performing network routing functions. In one example, an asynchronous transfer mode (ATM), Ethernet and other types of packets enter through the network interface MAC devices and are sent to the network processor <b>12</b>. These packets are processed in an application on the general purpose microprocessor <b>20</b> or on another processor that is coupled through the PCI bus interface (not shown). For reception and transmission of such packets, the application running on that processor <b>20</b> or the processor coupled through the PCI bus, makes use of a network stack <b>72</b>, which includes network management, control and signaling processes <b>74</b> to manage network communications.

0032The network stack <b>72</b> and the application run in the processor <b>20</b> that controls the microengines, or another processor coupled to the PCI bus. The paths of receive, transmit and data forwarding represent the transport of the packets through the processor <b>12</b>. The management control, signaling, and the network stack <b>72</b> usually are not involved in data forwarding. Essentially, the processor <b>20</b> receives and transmits. The processor <b>20</b> generates new packets that are transmitted over the network. The processor <b>20</b> can be involved in data forwarding in the exceptional case. This would involve very unusual packets, which may need special handling and complex processing.

0033For data forwarding processes, the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>are used. In some instances, data forwarding may occur at the general purpose processor <b>20</b> level. The signals Init is programmer's interface for initialization of microengine code. The signal Fini is used for termination (to put control info in a known state). The microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>provide fast, store and forward capabilities. The engines use a multilayer generic look-up process that performs validation, classification, policing and filtering using parallel hardware supported threads of the process. Exceptions and control packets are passed to the processor <b>20</b> for processing at the network stack <b>72</b>. A ternary network stack (not shown) can be located off-chip at a host via the PCI port or device port. This can be used to off-load the processor <b>20</b> or centralized management and control for one place. In some embodiments, the microengine is a compact RISC processor and can have limited instruction space. For this reason and for other reasons, it is desirable to reduce instruction code size when running multiple protocols. The network processor <b>12</b> implements a generic forwarding process that can be used to handle various protocol types (both existing and future types) without exceeding instruction storage limits.

0034Referring now to <figref idref="DRAWINGS">FIG. 4</figref>, an management arrangement <b>80</b> for forwarding table structures <b>90</b> that are stored in memory is shown. The forwarding table structure management <b>80</b> includes a control and management structure <b>82</b> including a network stack interface <b>84</b> and table managers <b>86</b>. The table managers <b>86</b> manage routing tables <b>90</b> that are stored in SRAM and can include a plurality of tables such as shown in <figref idref="DRAWINGS">FIG. 4</figref> including a layer <b>4</b> connection table <b>92</b>, a layer 3 destination table <b>94</b>, a layer 2 bridge table <b>96</b> and a layer 2 connection table <b>98</b>. Additionally, data structures stored in memory can include a packet buffer <b>100</b>, which is stored in DRAM. The microengines acting as packet data forwarding processors retrieve information from the routing tables <b>90</b> in SRAM and store and forward the packet information from the packet buffer in DRAM. The multiple tables <b>90</b> are set up by the control management processor <b>20</b>. For example, a layer 2 connection table <b>96</b> can be used for ATM virtual circuits, frame relay connections MPLS labels or other low level connections. A layer 2 bridge table <b>96</b> could be used for Ethernet bridging. A layer 3 destination table <b>94</b> could be used for Internet protocol (IP) forwarding based on a destination IP address. The layer 3 connection table <b>92</b> could be used for IP forwarding based on source and destination ports, addresses and protocol. All these tables may require that the packet be decapsulated or encapsulated.

0035Once the tables <b>90</b> are populated with forwarding information in a generally conventional manner, packet data forwarding processors can receive packets, perform table look-ups to obtain information and convert packets as required by the table entry. The control management process sets up the tables <b>90</b> with a common format for the purpose of decapsulation and encapsulation.

0036Referring now to <figref idref="DRAWINGS">FIG. 5</figref>, exemplary table entries, a subset of which are included in each of the tables <b>90</b>, is shown. The table entries include the following fields:

Forwarding Table Format

0000<ul id="ul0001" list-style="none"><li id="ul0001-0001" num="0000"><ul id="ul0002" list-style="none"><li id="ul0002-0001" num="0037">Decap Flag. Indicates whether the bytes should be stripped from the packet. If this flag is asserted, then the number of bytes to strip is in Decap Byte Count field.</li><li id="ul0002-0002" num="0038">Decap To Layer. This field specifies decapsulation of header layers up to the specified layer. The length of the layer and hence the decapsulation is determined by parsing the packet header.</li><li id="ul0002-0003" num="0039">Decap Byte Count. This field specifies the number of bytes to remove from the front of the packet. Decap is performed by adjusting the packet start offset in the packet buffer.</li><li id="ul0002-0004" num="0040">Current Encap. This field specifies an identifier of the current packet encapsulation type.</li><li id="ul0002-0005" num="0041">Encap Flag. Indicates whether bytes should be prepended to the packet. If this flag is asserted, then the number of bytes is in Encap Byte Count field, and the bytes to be encapsulated is in the Encap Header field.</li><li id="ul0002-0006" num="0042">Encap Byte Count. Number of bytes to be prepended to the packet.</li><li id="ul0002-0007" num="0043">Encap Header. The actual bytes to be prepended.</li><li id="ul0002-0008" num="0044">Next Table Type. If non-zero, this indicated that a further lookup is required. This gives the table type. For example, layer 3 routing or layer 3 connection table type. A layer 3 routing lookup would use a longest prefix match lookup algorithm using the destination IP address. A layer 3 connection lookup would use a 104 bit hash algorithm using source and destination addresses, source and destination ports, and protocol.</li><li id="ul0002-0009" num="0045">Next Table Addr. There can be multiple next tables, and multiple next tables of the same type. This field specifies the base address of the table.</li></ul></li></ul>

0046The flags get set or cleared by the management process. Signaling and setting up connections are part of the network system that will determine that a certain path through the network requires a change of the header. There can be many reasons why a header can change. Usually a header change is used when the protocol changes from one network domain to another.

0047Referring now to <figref idref="DRAWINGS">FIG. 6</figref>, a process <b>110</b> for encapsulating/decapsulating generic protocols is shown. Initially one of the microengines <b>22</b><i>a</i>–<b>22</b><i>f </i>receives <b>112</b> a packet from the network interface. The packet is comprised of one or more headers followed by a payload. The microengine, e.g., <b>22</b><i>a </i>copies the payload portion of the packet to a packet buffer in DRAM and it may place the packet at an offset in the buffer to make room for any new header that could be prepended to the packet for packet forwarding. The packet offset parameter for that packet is set to a default value determined at the offset into the buffer. The microengine reads <b>114</b> in the first header of the packet and performs a layer 2 look-up. The layer 2 look-up will read the table layer 2 bridge table and/or layer 2 connection table. The tables will return various parameters such as decap or encap flags. The process <b>110</b> will determine <b>116</b> if the decap or encap flags are set. If the decap and encap flags are set, the process will add <b>118</b> the decap byte count to the packet start offset and will subtract <b>120</b> the encap byte count from the packet start offset and prepend the encap bytes to the packet. The process <b>110</b> tests <b>122</b> if there is a next table to examine by looking at the blank field in the currently read table. If there is a next table, the process <b>110</b> will parse the next header <b>124</b>, fetch and read the next table. The process <b>110</b> continues looking to test the decap or encap flags being set.

0048If, however, the process did not determine that the decap and encap flags were set (<b>116</b>, above), it would determine <b>130</b> if the encap flag or the decap flag were set <b>132</b>. If the encap flag was set, it will subtract <b>120</b> the encap flag byte count from the start offset and prepend the encap bytes to the packet. On the other hand, if the decap flag was only set <b>132</b>, the process will add <b>134</b> a decap byte count to the buffer offset and, in any event, will check the next table <b>112</b>. When the process determines that it is at the end of checking the tables, it will then classify and forward <b>136</b> the packet in a conventional manner. That is, the “no” condition indicates that the process can classify and forward. Forwarding the header can have the microengine take the header and send it to the processor <b>20</b> or elsewhere, so that it can get reassembled with the payload. Forwarding the header could also involve forwarding the packet, etc.

0049Referring now to <figref idref="DRAWINGS">FIG. 7</figref>, in addition to specifying byte dissemination counts obtained from the look-up table. The look-up table may have the decap to layer field set in the table. This field specifies that the front portion of a packet should be decapsulated up to a certain layer. As known, packets are defined in protocol layers used in the OSI (Open Systems Interconnect) seven layer network protocol. After passing through physical layer 1, the first software layer seen by the network processor layer is layer 2, also referred to as the link layer. The length of the bytes to be decapsulated is determined by parsing the packet layers prior to the layer that is to be the new start of the packet. The length can be added to the packet start offset.

0050<figref idref="DRAWINGS">FIG. 7</figref> shows a variation where the decapsulation length is not specified in the table, but is determined by reading the packet itself. In other words, this would be a set of routines that would be inserted into the processing of <figref idref="DRAWINGS">FIG. 6</figref> substitute encapsulated byte count from the packet into the offset.

0051A process <b>140</b> to determine this offset is shown in <figref idref="DRAWINGS">FIG. 7</figref>. The process <b>140</b> includes reading the table <b>142</b>, determining that the decap to packet layer bit <b>144</b> has been set, and if set, retrieve the length of the layer to be removed by parsing the header <b>146</b> and adding the length to the packet start offset <b>148</b>. If the decap layer has not been set then the process simply skips. In any event, this process can be prepended to the process described in conjunction with <figref idref="DRAWINGS">FIG. 6</figref>.

0052A typical use of a decap to layer bit is to specify a decapsulation up to the layer 3 IP header. If the packet encapsulation is a multiprotocol over an ATM network such as the RFC 1483 standard, the layer 2 header length is determined by parsing the layer to header itself using the RFC 1483 length rules. However, if the packet encapsulation is classical IP the layer 2 length is determined by following the classical IP layer length rules. The packet encapsulation may be known by the port type it came in on from the prepended custom header from that port or may be obtained from the first look-up table in the current encap field.

0053Rather that each network protocol defining a separate protocol conversion this technique provides a generic approach. The approach saves code space and software development time-to-market. In an alternative embodiment, this technique can be implemented as a software library routine, e.g., a generic software building block for decapsulation/encapsulation, where customers can insert their proprietary header encapsulation and a customer's vendor need not get involved with customer's proprietary protocol designs.

0054A number of embodiments of the invention have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims.